In this document we outline our proposal for the renewal of our Center for the Multiscale Analysis of Genetic Networks (MAGNet). Over the last funding period ('05-'10), the MAGNet Center has made major progress in the description of molecular interactions involving proteins and DNA, in their functional analysis within specific cellular contexts, and in using this information to elucidate mechanisms controlling physiological and pathological phenotypes. As documented in this proposal, MAGNet has a compelling publication record, has made important discoveries across multiple scales of biological and disease related processes, and has developed key algorithms, models, and software that have been broadly adopted by the biomedical research community. We now plan fundamentally new research directions in multiple areas while, in parallel, achieving full integration of our core Structural Biology and Systems Biology themes. Indeed, as discussed in the proposal, these themes are highly synergistic and in combination can help dissect the relationship between atomic level changes (genetic variability) and cellular changes (phenotypic variability), with obvious applications to the elucidation of causal mechanisms in human disease. Center activities will involve a significant, multidisciplinary effort that will tackle multiscale problems, ranging from the atomic-level modeling of protein interaction specificity, to the reverse engineering of multi-layer regulatory networks, to using these models for the interpretation of the role of genetic variability in determining cellular phenotypes. These activities will directly impact three Driving Biological Projects aimed at (a) studying the DNA-binding specificity of key developmental transcription factors (Hox proteins), (b) modeling ErbB signaling pathways in oncogenic contexts using multi-factorial data and (c) assembling the first in vivo, genome-wide, regulatory network for prostate cancer using molecular profiles from chemical perturbation of human xenografts. While pursuing its tradition of scientific excellence, the center will continue to play a prominent role in the dissemination of the tools, models, data, and algorithms developed by its investigators. This will be accomplished primarily through geWorkbench, MAGNet's integrative bioinformatics platform, which has matured into a highly compelling and heavily used tool, as shown by its endorsement by caBIG and by its integration with other leading software tools such as GenePattern, Cytoscape, and Bioconductor. MAGNet will also play a key role in the continued development of our advanced data center, which provides our investigators with access to unique computational facilities and thus facilitates significant progress on research problems that would otherwise be inaccessible. Through MAGNet, we will further improve and extend the educational activities started in the previous funding period, which have produced a truly integrated experimental-computational curriculum. We will also explore a variety of options for dissemination of Center results and for the organization of community-based events, such as the now very successful DREAM and RECOMB Systems Biology conferences. Finally, MAGNet has played a central role in the development of an inter-disciplinary program in Computational Biology at Columbia University that spans two campuses and seven academic departments. We believe that the unique research environment we have created can serve as a model for the full integration of Computational Biology in all areas of biomedical research.